Re: CHAT: Phonemic status of English interdentals

Adrian Morgan wrote:

>Tristan wrote:
>
>
>
>>Much easier to tell the difference between [&] and [&:] (which only
>>have one debatable minimal pair---can and can)
>>
>>
>
>1. "banner" (one who bans) vs "banner" (flag)
>
>2. "banning" (participle of ban) / "Banning" (surname of former South
>   Australian premier.
>
>
They can, apparently, be excused on the grounds of having a morpheme
boundary in them.

>>or [8u] and [Ou]
>>
>>
>I don't understand the rationale behind your choice of symbols. If [8u]
>is supposed to represent the diphthong in "ode", then surely [3\}] is
>a better transcription). [8] being higher than schwa and [3\] being
>lower, but both being central rounded vowels. I use [8] to denote the
>vowel in "bird".
>
You could be right. I use [8:] to denote the vowel in 'bird'. The first
element of the vowel is almost a rounded backish [@] and so neither
really closer to [8] nor [3\], so I use [8] for simplicity. I think. I
might actually be horribly wrong. Can argue simplicity anyway, because I
wasn't being especially narrow in my transcription? (I used [u] to
indicate a vowel closer to [}], after all.)

>And there is a minimal pair - ode [3\}d] vs old [Oud], unless you
>speak a dialect in which /l/ is not reduced to [u] in this position.
>
And indeed I don't. That is, I pronounce /l/ in that position as a nice
and dark /l/, but an /l/ nevertheless.

Tristan